Stop pretending TTRPGs have objective difficulty.

[Koumei]

While the CR system works as a guideline for MCs ("Okay, they're level 6, let's see what a good challenge might be... Dire Otter, that's CR 6. Okay, now let's look over the statline and see if it is in fact appropriate" as opposed to "Okay, they're level 6, now let's look over the stats of every monster ever until I find something appropriate"), I think it works best as a benchmark system for players: it tells them how tall they should be at any given level.

Which is the Same Game test, basically, but on a per-character basis where you have actual numbers on your sheet there rather than "Probably has a +9 to this and would have a couple of spells of X level, let's say Charm Monster and Black Tentacles."

[Fuchs]

Yeah, but the same game test requires the same GM. You can't use CR as a baseline to judge whether or not campaign A is harder than campaign B when those are not done by the same DMs.

[hogarth]

Yeah, but the Same Game Test requires the same GM.

Even using the same GM isn't good enough for the Same Game Test, because both the player and the GM have the capability to either try hard or sandbag depending on what they want the SGT to say.

Where the CR system does a useful job (IMO) is having a chart that says "a level N creature has roughly a +X attack bonus, Y hit points and armor class Z"; that may be either a huge challenge or a speed bump to a player with a level N character in D&D (since so much depends on tactics, equipment, feat choices, etc.), but at least it's easy to eyeball whether a particular PC is unhittable or likely to be killed in one shot.

[Murtak]

Godfuckingdamnit, where does this "but different DMs are different"-argument come from? Why should it matter that different DMs do things differently? Of course they do! So what? We flat-out do not care about cross-campaign adjustments. Guess what, the same fucking DM is different from day to day, from mood to mood and from player to player. Some go easy on newbies, some go easy on their girlfriends/boyfriends and some make the final fight challenging on purpose. But that does not mean challenge ratings (or similar systems) do not matter. You could even argue they are all the more important because of these disparities. Having all your monster, traps and puzzles lined up on a sliderule of difficulties is immensely useful for both DMs and players to be able to adjust the difficulty of their encounters. And no matter how much DMs and players vary, this is useful knowledge! When someone puts a third level party up against a Glabrezu and it eats them the players can see their DM is an idiot and the DM can see both that he should have used a Babau instead and what other, easier monsters there are to use. Properly populated you can use challenge ratings to build worlds, campaigns and encounters and you can even use it to check your party members on whether they are up to par. You can even use it to adjust difficulty, by using different encounters as a DM or by not visiting the Crater of demonic Doom as the players.

And all of this applies no matter how different DMs and players vary from one game to another. It even applies to intentionally sandbagging. Heck it helps you in doing so.

[Fuchs]

This is about people claiming that they are objectively better players for defeating objectively harder challenges. CR can be a tool for DMing and eeballing ecounter difficulties, but it's not an objective tool to comapre two campaigns' difficulty rating.

[Username17]

The Chicago Bulls sometimes play really well and sometimes play really shitty - relative to what we expect from the Chicago Bulls. But despite the fact that they don't put the same numbers of the scoreboard every night, I doubt many people would be making the incredibly bizarre claim that the Chicago Bulls are not objectively stronger than the Arlington Highschool Basketball Team.

Objective criteria exist even when there are a lot of variables. Heck, in many cases the existence of a lot of variables makes the objective criteria more important, since islands are more needed in stormy seas.

-Username17

[Fuchs]

The Chicago Bulls sometimes play really well and sometimes play really shitty - relative to what we expect from the Chicago Bulls. But despite the fact that they don't put the same numbers of the scoreboard every night, I doubt many people would be making the incredibly bizarre claim that the Chicago Bulls are not objectively stronger than the Arlington Highschool Basketball Team.

Not many people would make the incredibly bizarre claim that a balor that wades into melee is equally difficult to beat as a balor that teleports away and starts a hit and run fight using all its abilities, especially summoning more fiends.

Your example is shit. More fitting would be comparing the Arlington Highschool basketball team to the Arlington Highschool female soccer team. Yes, both are sports teams, but they are not playing the same opponents, they are not even the same game.

That's how much of a difference a DM makes.

[crasskris]

Objective criteria exist even when there are a lot of variables. Heck, in many cases the existence of a lot of variables makes the objective criteria more important, since islands are more needed in stormy seas.

But even objective criteria have a specific target, a concept they are supposed to measure. Using the scoreboard number to infer the dribbling ability of a team would be sketchy, at best.

And so is using CR as an (objective) measurement for the abilities of players. A rough indicator, perhaps. But not an objective measure.

[hogarth]

The Chicago Bulls sometimes play really well and sometimes play really shitty - relative to what we expect from the Chicago Bulls. But despite the fact that they don't put the same numbers of the scoreboard every night, I doubt many people would be making the incredibly bizarre claim that the Chicago Bulls are not objectively stronger than the Arlington Highschool Basketball Team.

But I think the point is that saying "Red Hand of Doom was totally easy for our party" is relatively meaningless because the DM could be playing the bad guys like the L.A. Lakers or he could be playing them like West Arlington Prep.

Of course, the best basketball comparison for D&D is the Harlem Globetrotters vs. the Washington Generals. How tough are the Washington Generals? It's hard to say objectively, when most of their games are spent losing to the Globetrotters.

There are certainly objective benchmarks that you can point at. E.g., if your fighter spends most of his time stabbing people with a sword, then it's meaningful to point at a line on the character sheet that says "+9 attack bonus, 1d8+7 damage". But "we beat up a dragon!" isn't one of them.

[Archmage]

This is undiluted 4rry bullshit. The game world is supposed to feel like a world, and it's supposed to have some fucking consistency. If you go to a police station, there are cops there. If you go to a military base, there are soldiers there. If you teleport to the Abyss, there are demons. Blah blah blah.

Don't pretend that this is true for any reason above and beyond "the MC put them there."

It doesn't offend you that "the MC put them there" when you go to Baator and randomly encounter a balor but it does feel like total bullshit when you walk down the road outside a rural hamlet and a balor teleports into the party's path is because one of these things seems more logical; after all, there are balors in hell, so you might meet one if you go to hell. But there's nothing stopping a balor that happens to be on the material plane from teleporting into your path on a country road. I agree with you completely that player decision-making determines how "difficult" the game is going to be. Obviously, players who do things like pick a fight at the police station or plane shift to Baator are going to a place where encounters will be more likely to go poorly.

But unless you literally determine everything that happens in a particular game with die rolls from unbiased tables generated without foreknowledge of the PCs or even the tactics commonly used by specific players, "the police station is full of police because police stations are full of police" is a meaningless statement as far as gauging objective difficulty. Do the police have six-shooters or riot shotguns? Are they using gas grenades or other crowd-control measures effectively? How many cops are actually in the building at a given time? How long does it take a first-responder team to show up with armor and assault weapons? How tactically-competent are the local police, anyway? How hard it is to conquer the police station is completely dependent on how the MC designed the police station. The fact that the players can case the joint and learn all of these things before they decide whether or not to attack the police station does not change the fact that the MC probably didn't design the police station in a vacuum. And odds are that if the players have decided on a whim "we're going to assault the police station to retrieve the evidence they gathered against us earlier" that the MC doesn't even have a pre-written and pre-populated "police station dungeon" and is therefore going to have to pull the whole sidequest out of his ass.

The fact is that the "game" aspect of TTRPGs is bullshit covered in more bullshit hidden behind a smokescreen that makes it look like the MC isn't letting you win. Because every time you "win," it's because the MC let you win by not putting a balor in the police station when he is well-within his rights to decide that there's a fucking balor in the police station. As a player you can call bullshit ("why is there a fucking balor in the police station") and walk away from the table if you don't think you're being treated fairly, and it makes you a shitty MC to spring surprise balors on PCs when they have no chance of winning, but the fact that the world is supposed to feel like a real world with consistency does not change the fact that no matter where the PCs encounter a balor the MC fucking put it there and every time they encounter a group of bandits they can conceivably best in combat the MC is responsible for putting those there, too.

[Ektagliaresia]

The only way you can create "objectively difficult" challenges is to write a module blind with no knowledge of the PCs that will be played in that campaign. Which I think most people are going to agree results in shitty railroaded games because at that point you have by definition not allowed yourself to improvise new content.

What you just described is a sandbox campaign, the opposite of a railroad. Making material in response to the PCs is different from improvising new content; in the same way that I can improvise an adventure right now, I can improvise the same adventure during play.

In a full treatment sandbox game, the MC doesn't write a single module, he writes a game world large enough to accommodate a campaign in which the PCs act, filling in the blanks with random tables. This creates a campaign of objective difficulty, even if that difficulty is somewhat variable.

But unless you literally determine everything that happens in a particular game with die rolls from unbiased tables generated without foreknowledge of the PCs or even the tactics commonly used by specific players, "the police station is full of police because police stations are full of police" is a meaningless statement as far as gauging objective difficulty. Do the police have six-shooters or riot shotguns? Are they using gas grenades or other crowd-control measures effectively? How many cops are actually in the building at a given time? How long does it take a first-responder team to show up with armor and assault weapons? How tactically-competent are the local police, anyway? How hard it is to conquer the police station is completely dependent on how the MC designed the police station. The fact that the players can case the joint and learn all of these things before they decide whether or not to attack the police station does not change the fact that the MC probably didn't design the police station in a vacuum. And odds are that if the players have decided on a whim "we're going to assault the police station to retrieve the evidence they gathered against us earlier" that the MC doesn't even have a pre-written and pre-populated "police station dungeon" and is therefore going to have to pull the whole sidequest out of his ass.

Some parts of the game world being filled in after the fact doesn't make the whole game into "hope the MC lets us live" unless the MC is basing his improvisations on the PCs. Which he doesn't have to do.

[crasskris]

This creates a campaign of objective difficulty, even if that difficulty is somewhat variable.

Y...no. This may create a campaign that allows for some objective criteria that serve as indicators. Any interpretation of those criteria is subjective (although subjectiveness may be based strongly on experience or common sense) interpretation. Actual in-game difficulty again rests largely on the shoulders of MC and players alike.

Some parts of the game world being filled in after the fact doesn't make the whole game into "hope the MC lets us live" unless the MC is basing his improvisations on the PCs. Which he doesn't have to do.

Can you explain that one further? As far as I see it, the MC always has some wriggling space in his improvisations, which might (consciously or unconsciously) affect difficulty.

[infected slut princess]

Does it, or does it not make sense to complain about Giant Crabs?

[shadzar]

Some parts of the game world being filled in after the fact doesn't make the whole game into "hope the MC lets us live" unless the MC is basing his improvisations on the PCs. Which he doesn't have to do.

Can you explain that one further? As far as I see it, the MC always has some wriggling space in his improvisations, which might (consciously or unconsciously) affect difficulty.

take a DM written adventure and something that has been there BEFORE this group of players. this isnt an "officially published" adventure, but it exist prior to the players.

Most DMs wont rewrite all their work just because the current group varies from the last one that ran through the adventure. this is the EXACT same as a published adventure.

so a part added because the current group take the path to the left that the previous group didnt, so isnt fully fleshed out means the adventure might be somewhat tailed to the current group for that left path, while the path to the right will remain the same.

likewise the 3rd group upon choosing either path will have group 1's right hand path, and group 2's left hand path, neither of which were designed with group 3 in mind, even if somewhere else or along the way some encounter was created JUST FOR group 3 that fits within their current level.

maybe group 3 decided to evade an encounter that the previous groups didnt so are at a lower level and an encounter was created for them before getting to the fork in the road that leads to the group 1 side, or group 2 side.

NOT everything is created on the fly, so predesigned things are there and ready to go so the gameplay doesnt slow down, and like adventures they get reused independent of the current groups stat.

Encounter Options
Once an encounter occurs, there is no set sequence for what happens next. It all depends on just what your characters have encountered and what they choose to do. That's the excitement of a role-playing game--once you meet something, almost anything could happen. There are some fairly common results of encounters, however.
Evasion: Sometimes all you want is for your characters to avoid, escape, or otherwise get away from whatever it is you've met. Usually this is because you realize your group is seriously outmatched.

Copyright 1999 TSR Inc.

not everything is there because it is level appropriate. sometehing video games adopted form D&D. placing things in places where the THING belong, not just because the players are in a level 1 town.

organic v inorganic play.

the world exists outside of the location the players are in.

the PHB does NOT contradict the DMG...

The Encounter is Too Difficult

The DM has accidentally pitted his player characters against a group of creatures too powerful for them, so much so that the player characters are doomed.

Copyright 1999 TSR Inc.

in some cases the DM places things where THEY belong, so the players deciding to fight it rather than run are and should be, allowed to have their characters die. the DM making an encounter that should be within range of the players, but screwing it up due to bad design, is the ones the DM fixes.


GOOD:
DM: oops i thought you could take Tucker's Kobolds at your level, let me fix this

PCs: we aren't ready to fight a next of dragons, lets go somewhere else.

BAD:
DM: its Tucker's Kobold so you are stuck fighting all 2000 of them.

PCs: alright we get to beat a whole nest of dragons at our level!

the DM failed to realize and fix his mistake, and the players failed to understand everything isnt there for them to fight RIGHT NOW (when they first encounter it).

again playstyle issues and the way one views the game becomes a problem.

[Ektagliaresia]

This creates a campaign of objective difficulty, even if that difficulty is somewhat variable.

Y...no. This may create a campaign that allows for some objective criteria that serve as indicators. Any interpretation of those criteria is subjective (although subjectiveness may be based strongly on experience or common sense) interpretation. Actual in-game difficulty again rests largely on the shoulders of MC and players alike.

Admittedly, the fact that the MC is not a computer prevents true objective difficulty. However, this kind of objectivity is ideal for a sandbox game. If the MC were replaced with a computer, the campaign would have a truly objective difficulty, but, of course, tabletop RPGs have MCs with Rule Zero to create games in which players can improvise beyond programmed limits. So, a campaign can have an objective difficulty insofar as anything in an RPG can be called "objective".

Some parts of the game world being filled in after the fact doesn't make the whole game into "hope the MC lets us live" unless the MC is basing his improvisations on the PCs. Which he doesn't have to do.

Can you explain that one further? As far as I see it, the MC always has some wriggling space in his improvisations, which might (consciously or unconsciously) affect difficulty.

Once again, I admit that knowledge of the players and their PCs will inevitably taint improvisations by the MC. However, this is different than the MC willfully basing his improvisations on the thoughts and desires of the players or the abilities of the PCs. For example, MC has not mapped out the interior of the lord's manor, and the PCs sneak into it, the MC could think "what kind of layout and contents would likely be in this manor?" or "what layout and contents would be challenging but not overwhelming for the current party?".

[crasskris]

For example, MC has not mapped out the interior of the lord's manor, and the PCs sneak into it, the MC could think "what kind of layout and contents would likely be in this manor?" or "what layout and contents would be challenging but not overwhelming for the current party?".

Well, in this example it may be perfectly reasonable to assume the lord employs a wizard, and just as reasonable to assume he doesn't. But the existence might greatly influence the difficulty encountered, especially in 3e-.



My point is: I don't think an objective assessment of difficulty is possible a priori, just from scanning the adventure layout. Just as the stat block and common tactics section of the Balor description provide only a difficulty baseline, a static description of elements planned for an adventure can only give you an rough indication of what difficulty the adventure might provide. Which serves the function of a guideline for MCs to pick adventures well, but fails as an useful, objective measurement of player 'skill'.

I do, however, think there is objective difficulty in TTRPG, or at least a workable approximation of it.
As Lago implicitly demonstrates this can only be measured afterwards - by a comparison of technical prowess (i.e. the abilities and numbers) of both encounters and players, and by judging the tactical savvy the MC had shown (provided the judge is experienced and has demonstrated common sense at some point). In this scenario the raw, written description of the adventure only serves as a template, both for play and for communication of the results.

So yeah, I agree with the assessment that "We beat Adventure XYZ without breaking a sweat" is anecdotal at best and holds only an indication of actual skill involved. Without a detailed recounting any useful objective measurement is doomed to fail.

[Dean]

Of course, the best basketball comparison for D&D is the Harlem Globetrotters vs. the Washington Generals. How tough are the Washington Generals? It's hard to say objectively, when most of their games are spent losing to the Globetrotters.

I think this is very well put. You have been "Quoted for Truth" Sir.

[Murtak]

So yeah, I agree with the assessment that "We beat Adventure XYZ without breaking a sweat" is anecdotal at best and holds only an indication of actual skill involved.

But even that is useful information. Given that +2CR = double the challenge (which holds up fairly well in my opinion) you need just a couple of data point to get a pretty good idea of hard the encounter actually is. And even a single data point will give you a decent hint. It may not be an accurate measurement, but that does not mean it is random or useless.

[crasskris]

Given that +2CR = double the challenge (which holds up fairly well in my opinion) you need just a couple of data point to get a pretty good idea of hard the encounter actually is. And even a single data point will give you a decent hint. It may not be an accurate measurement, but that does not mean it is random or useless.

As I said, it is an indication that has its uses.
But an adventure is a string of encounters, with any number of possibly CR-less interlude challenges. This adds a level of complexity and flexibility that, imho, makes an a-priori measurement of actual difficulty extremely hard, since all you have for data is a couple of CR values and a more or less informed assumption of probable player actions.

[Murtak]

Those assumptions and CR rating have been extremely useful and - with the exception of some outliers - been decently accurate. As far I can tell even the places where CR was clearly off were just instances of wrongly assigned CR and not the system itself being at fault. As far as I can tell 3Es CR system works.